in ADL 1.5 templates, the language handling is the same as for current
archetypes. And it is certainly not the case that English (let alone UK
English) is the original language of all archetypes. We already have
archetypes whose original language is PT-BR, NL, and (I think) Russian.
- thomas
Hello List members:
(mucho gusto David)
The translations of archetypes were part of the final evaluation of Ing. Pablo Pazos first online course of openEHR in Spanish.
I was rather shocked to see so few archetypes in Spanish, so hands on task I´ll translate myself, as many as possible. So I started with archetypes in English. I found previously David’s translation and I think there is another from Chile.
I looked at both; they look fine to me, so I didn’t make any variations.
I believe Ian´s approach to using a root language is correct. I couldn´t find any impediment to use my es-ar translation for es-es (for the body temperature archetype).
I know a fellow classmate Leonardo will also be translating archetypes, we will work together and not overlap.
I do have a suggestion that may be useful in this discussion: Our School of Medicine at the UM works extensively with a Languages Lab (many languages even oriental), so I can do a small research about the variability in different variations of languages and if it impacts the meaning of a clinical concept.
I hope to be able to tap this resource into the openEHR community somehow too.
Hope this can be helpful
Best regards
(PS: When I´m knowledgeable enough to create archetypes myself they will be done in both Spanish and English)
Dr. Domingo Liotta (h)
Director Cátedra Abierta
de Bienestar Médica, Universidad de Morón
My profiles: 
Contact me: 
domingo.liotta 
dliotta 
domingo_liotta@hotmail.com 
domingo_liotta@yahoo.com.ar
| My latest G+: Facebook Off (preparando proyectos para el 2012) Saludos!!
My G+ - Posts - Add to Circles - 14:34 Dec-11
|
- | - |
| Latest tweet: @valelarravide hola Vale
si porsupuesto Mandame un DM tranqui
Follow @dliottaReply
Retweet %20si%20porsupuesto%20Mandame%20un%20DM%20tranqui)23:11 Jan-12
|
- | - |
I accept that there are several languages, now.
But when it comes to connecting e uniform meaning to these strings of text we need to define the Golden Standard, so we all know what is meant.
Or do you accept as expression of the meaning a string of text you do not understand?
Gerard Freriks
+31 620347088
gfrer@luna.nl
Hi Gerard,
The gold standard in openEHR archetypes is always the primary authored
language. In practice many non-English-authoring authors will choose
to provide an English translation but it is ultimately up to the
consumer of the archetype to ensure that translations (to any
language) are correct.
Ian
Dr Ian McNicoll
office +44 (0)1536 414 994
fax +44 (0)1536 516317
mobile +44 (0)775 209 7859
skype ianmcnicoll
ian.mcnicoll@oceaninformatics.com
Clinical Modelling Consultant, Ocean Informatics, UK
Director/Clinical Knowledge Editor openEHR Foundation www.openehr.org/knowledge
Honorary Senior Research Associate, CHIME, UCL
SCIMP Working Group, NHS Scotland
BCS Primary Health Care www.phcsg.org
I understand.
But this is not the way the create a uniform meaning.
Gerard Freriks
+31 620347088
gfrer@luna.nl
I don’t dispute that point, but the problem is the same no matter which language is the original language for a given archetype - it is just that different people do or maybe don’t understand. If the editorial group for a particular archetype happens to be a bunch of mainly Russian docs, then the problem is not them understanding the archetype, but others may. I suppose we could require that EITHER an archetype is originally authored in English OR it is translated to English before publication, but I would be uncomfortable with that. It doesn’t fit the reality of South America, eastern (Russian speaking) Europe, and other specific regions in the world - each of these regions, when sharing documents with each other tend to use the highest fidelity common language available to them - whcih could be Spanish, Russian, etc etc (it might also be English but I think there is no guarantee).
well that's true, but from an IT point of view, creating 'meaning' isn't
done with language, it is done by placing concept nodes within an
ontology in correct relation to other nodes - the only 'true meaning'
is: how does reality classify against an ontology. Doing this with
archetypes requires binding concept codes from SCT and/or other
ontologies, where available. Which is certainly a challenge we face, but
it's not the same at the translation issue.
- thomas
Whatever people do locally, is their choice.
As EN13606 Association, when we will publish artefacts, the meaning of things inside are either coded using a Reference Terminology and/or a list with preferred and well defined names, under the control of one editing group.
Inside artefacts that are published for general use we must not allow meaning creep because of translation/language problems.
Gerard Freriks
+31 620347088
gfrer@luna.nl
Semantics isn’t about reference or preferred or well defined names pre - established. It’s about how can my system can understand the meaning of your system without any previous agreement. What did you mean and not what you said. Thesauri help us to understand each other, but only if i can understand the context. The main challenge of information systems is to link environmnt to language. Environment is the info model, theontology of a certain domain. Terminology is more universal, can be the língua franca , but it must be referred to a particular context.
I totally agree with Ian here...The gold standard should be the original language - whatever that is. And it is the responsibility of the people who are using it clinically (even on paper form) to make sure the meaning is not altered. I think the same argument would apply to Merck Manual which is originally published in English (us) but has translations in almost any language on which many clinicians depend on when making decisions... QA by a reputable body (in this case CKM editorial group) is essential.
Terminology is always good but not essential and I suspect there may be conflicting ontological sources out there (e.g. describing same concepts in a different way).
Cheers,
-koray
I don't think there is any doubt that the original language is 'king' -
after all in captures the intent of the authors best.
Translations are ideally added after the archetype is stable and has been
published. We saw a perfect example of how translation efforts can go to
waste when an archetype is translated while still in draft, where all that
was safely salvageable of the German and Arabic translations for the Adverse
Reaction archetype were a few nodes when the content was significantly
changed.
However also keep in mind that while we are collecting quite a number of
translations for both draft and published archetypes, we have not yet had
any translations verified - they are only the work of an individual so far.
So part of our ongoing process will be to organise reviews of each
translation to ensure that the translation is also true and safe.
Similar processes are in place in CKM for terminology binding reviews - to
enable those with terminology insights to ensure that where they are bound
in an archetype, they are the best for the job!
However the original question Ian raised was around how to be more efficient
in managing our translations and I'm not sure that there is an easy answer
here. Translators will hopefully be able to guide those of us severely
limited to English. Domingo Liotta translated an archetype recently into
Argentinian Spanish and when asked he confirmed that there was nothing
specific to the Argentinian translation that was not appropriate for the
Spanish parent, so this single translation has been uploaded for both. This
is a bit clunky but probably the way we will have to work in future,
gradually understanding where we might leverage strong language similarities
and where this will not be possible. Whether we can upload a translation for
a parent language and update certain nodes in a specific child language does
not necessarily appear to be a less complex operation. We will learn, no
doubt!
Cheers
Heather
Hi Heather and all,
I think translation is only one of the alternative concepts for the original.
Validation for translated artefacts is much controversial, because there are
many answers for it and no perfect answer especially in subjective matters.
Terminology system provides most possible candidate in semantics, but it is
limited as Thomas Beale said.
Each localisation committee could be engaged in translation management
to add some proof of validity. What about such community proof for?
Best regards,
Shinji Kobayashi
I think it an excellent idea that the localisation committees take on
responsibility for initial translations and subsequent validation of
translation thru reviews. Always happy to delegate.
Let me know when Japan is ready to go:)
And in the case of families of languages, perhaps these localisation
committees can collaborate to solve our translation burden - to seek
efficiencies instead of everyone reinventing the wheel for each translation.
Regards
Heather
Funny, that coincidentally IHTSDO sent today an email recruiting someone of its community to a new job position of content change coordinator, someone to manage the end to end process of any change of snomed ct, including mapping to Spanish and other languages and special country extensions. Quality assurance of mappings and translations are crucial to the success of any federated initiative like ours. IHTSDO is an example how hard can it be to work in a distributed, multilingual and multicultural environment. They have been developing tools and lots, lots of policies to make of Snomed ct a universal clinical reference terminology. So far, everybody knows, they are still far from it and the organization faces a lot of criticism because of the poor outcomes.
To delegate the governance of archetype translations to the localization committee is really a good idea, Shinji, but again, the worst is not to translate, it's to maintain. Without a sustainable business model, that assures resources to us to have someone, and not a voluntary committee, in charge of managing the whole process, I'm afraid this process could lack on continuity and quality.
I am the meaning (and I think Koray too), that the localization committee first task is to seek for sponsors for our national chapter. Long-term sustainability is my main concern. I don't think we should already commit to take over ( huge) tasks like that. Not before we have governance of all process.
There are some many issues that have arisen since we began to discuss the new business model and federation issues, that I think our unconference late this year should be take place earlier. Governance and not technical aspects are for us, countries that want to adopt openEHR as the national reference, the main issue we'd like that the localization committee could discuss now, because all the other things depend on it.
Regards
Jussara Rötzsch
Director, OpenEHR Foundation
Hi Jussara,
Yes, sustainability is important to keep quality. However, every big
open source software
project started in personal or small voluntary community. I do not
agree to raise up local
community requirements, because no one has enough resources to do so.
Ultimately,
government or NPO/NGO would be obliged to assure the quality, but
government do not
work without proven merits on it. I am talking about feasibility to
make an NPO to maintain
openEHR.jp within our Japanese community and we will have an
pre-meeting in these weeks.
I am an optimist and an evangelist of open source software movement.
If someone think
our artefacts lack something, he/she would join us to fix it like
Wikipedia. In my experience
on open source software community, good projects have power to attract
to manage, and
I also believe openEHR is enough good to collect people.
The ideal proof would be 'government proof' or 'government authored
committee poof'.
'Community proof' could provide some indication to decision for
developers and clinicians.
Actually, there are not so many Japanese translated archetypes, I am
not afraid to maintain
them. I would be very happy to fill up my capacity to deal such proof.
Best regards,
Shinji Kobayashi
Hi Shinji
I´m glad to find other evangelists of OpenEHR cause! I´m doing that for past eight years in Brazil, and apparently it has worked because Brazil is officialy adopting both archetypes as aslo OpenEHR RM officialy in our national ehealth program. But just because of that my responsiblility has grown exponentially and I have to make sure that OpenEHR will be a stable, sustainable and business driven organization. Pleople who criticize this option (not only HL7ers) says that to adopt openEHR is risky, because it´s a voluntary organization who hasn´t support of big players (Other open sources initiatives are supported by big companies, like Oracle for instance, or even by government, like here in Brazil, where the government has an opensource policy). The decision of the foundation to change its business model was among other things to raise support for our evangelists from the communities where they pray…
It´s not possible for one person that isn´t supported by his employer or by academia to dedicate her/his time to do, for instance, quality assurance. You know that the main reason we have so few archetypes already published it´s because only a few people are de facto commited to this task. I think the quality assurance should be one job of the operational group that will do paid work for the foundation. That means a work for the foundation and not for a committee. Like Apache, we must have get support and who else besides government or big comapnies can do that?
Once again I´m using IHTSDO, which is totally supported by governments as an example. It cannot be a perfect one, but it´s better than every other SDO (which, as a matter of fact also live of government grants, don´t they)
Regards
Jussara
Hi Jussara,
In Japan, the government have also appealed to promote open source software,
but they say OSS would not be applied because of lack of big company supprot
by the same mouth, too. They have tendency to believe US big company rather than
Japanese serious community.
Otherwise, only I can do is to publish my artefacts with open source software
condition to be available for everyone and I think I have responsibility to my
artefacts under Apache 2 license condition. If I thought I have
responsibility to
'harakiri' on my aretefacts, I cannot do anything and others, too.
I would take over this the next Japanese pre-meeting. Big company support is
very attractive, but sometimes it is beginning of death of community.
Best regards,
Shinji
One thing worth mentioning, it is quite a simple matter translating an archetype compared to all of the terms as words or phrases (even fully specified words) as words or phrases are used in many contexts.
Cheers, Sam
